The ansamitocin family of ansamycin-type microbial metabolites produced by the Actinomycete Actinosynnema pretiosum and the structurally almost identical plant-derived maytansine and its congeners are extraordinarily potent antitumor agents. Despite its toxicity, maytansine at one time was considered an outstanding candidate for clinical development, but it failed to show significant efficacy in phase II clinical trials, probably due to dose-limiting toxicity. However, interest in these compounds continues, e.g., as "warheads" for antibody-targeted delivery, and their high potency calls for further efforts to modify their structures with the goal of identifying analogs which retain high antitumor activity coupled with lower toxicity than the parent compounds. The structural complexity of these compounds limits chemical approaches to modified structures to those accessible by semi-synthesis from the natural product starting materials; more deep-seated backbone structural modifications require biochemical approaches based on genetic alteration of the biosynthetic machinery generating the parent compounds. With this rationale in mind we have cloned and sequenced the ansmitocin (asm) biosynthetic gene cluster from A. pretiosum and are in the process of analyzing the functions of its individual genes. In the next 4-year period of this grant we wish to continue this work by determining the functions of all the downstream modification genes/enzymes, expressing the polyketide synthase (PKS) assembling the backbone of ansamitocin in a heterologous host and studying its structure and mode of operation, and clarifying the structure and mode of formation of a rare polyketide chain extension unit required for the function of the asm PKS. Based on the insights gained from this work, we will then assemble a system, which allows the expression of all the asm biosynthetic genes to produce ansamitocins in a heterologous host, Streptomyces coelicolor, from a series of gene cassettes under the control of an external promoter. Once this system is established, we will demonstrate the feasibility of producing ansamitocin analogs by introducing genetic modifications into the expressed gene cluster. In addition we plan to investigate the genetic control of ansamitocin production in A. pretiosum with the aim to increase yields of the parent and engineered mutant compounds. This work will thus provide the tools for the preparation of structural analogs of the ansamitocins, which can be evaluated for an improved therapeutic ratio as anticancer agents.